Research of illumination automatic correction of face color image

摘要

Different environment illumination has a great impact on face detection and recognition. The automatic detection and radiation correction of a highlighted face area is helpful to analyze and identify human faces correctly in a color image. In this paper, the part of high-light and shadow of face skin color image was researched for reducing the impact of high-light and shadow for image processing. The high-light region was detected based on human face color and pixel spatial distribution characteristics for the high-light area. According to offset rate of each component in the rgb color space high-light field was corrected. Used edge fitting method corrected image in high-light area where the image over the display range. For the shaded area Retinex methods used to enhance intensity. Experimental results show that the approach of this paper proposed can be better to remove the impact of high-light and shadow.